Goldman Sachs chief economist Jan Hatzius suggests that the surge in AI-related capital expenditure is approaching its peak. Currently, US spending on AI hardware, data centers, and chips is estimated at approximately $600 billion, a figure expected to plateau in the future.
The contribution of AI to US GDP growth remains modest, recorded at 0.1 percentage points. Hatzius attributes this partly to the import of hardware, which causes a significant portion of spending to leak out of domestic GDP figures. He notes that the transition to a fully realized AI economy is a process spanning decades.
Apollo chief economist Torsten Slok reports that AI has provided zero financial payoff for companies outside the technology sector. While the Magnificent 7 firms have seen increases in profit margins and revenue per worker, these metrics remain flat for the remaining companies in the S&P 500 index.
Slok observes that many corporations now treat spending on AI tokens as a cost center rather than a growth investment. This creates a concentration risk for investors, as the economic benefits of AI appear to be limited to a small group of mega-cap companies rather than the broader market.
